ANSYS Mechanical APDL Multibody Analysis Guide 18.2
ANSYS Mechanical APDL Multibody Analysis Guide 18.2
ANSYS Mechanical APDL Multibody Analysis Guide 18.2
Guide
ANSYS, ANSYS Workbench, AUTODYN, CFX, FLUENT and any and all ANSYS, Inc. brand, product, service and feature
names, logos and slogans are registered trademarks or trademarks of ANSYS, Inc. or its subsidiaries located in the
United States or other countries. ICEM CFD is a trademark used by ANSYS, Inc. under license. CFX is a trademark
of Sony Corporation in Japan. All other brand, product, service and feature names or trademarks are the property
of their respective owners. FLEXlm and FLEXnet are trademarks of Flexera Software LLC.
Disclaimer Notice
THIS ANSYS SOFTWARE PRODUCT AND PROGRAM DOCUMENTATION INCLUDE TRADE SECRETS AND ARE CONFID-
ENTIAL AND PROPRIETARY PRODUCTS OF ANSYS, INC., ITS SUBSIDIARIES, OR LICENSORS. The software products
and documentation are furnished by ANSYS, Inc., its subsidiaries, or affiliates under a software license agreement
that contains provisions concerning non-disclosure, copying, length and nature of use, compliance with exporting
laws, warranties, disclaimers, limitations of liability, and remedies, and other provisions. The software products
and documentation may be used, disclosed, transferred, or copied only in accordance with the terms and conditions
of that software license agreement.
ANSYS, Inc. and ANSYS Europe, Ltd. are UL registered ISO 9001: 2008 companies.
For U.S. Government users, except as specifically granted by the ANSYS, Inc. software license agreement, the use,
duplication, or disclosure by the United States Government is subject to restrictions stated in the ANSYS, Inc.
software license agreement and FAR 12.212 (for non-DOD licenses).
Third-Party Software
See the legal information in the product help files for the complete Legal Notice for ANSYS proprietary software
and third-party software. If you are unable to access the Legal Notice, contact ANSYS, Inc.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. iii
Multibody Analysis Guide
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
iv of ANSYS, Inc. and its subsidiaries and affiliates.
List of Figures
2.1. FE Slider-Crank Mechanism .................................................................................................................... 5
2.2. Rigid Body Definition With Underlying Elements ..................................................................................... 7
2.3. Rigid Body Definition Without Underlying Elements ................................................................................ 7
2.4. Rigid Body with a Limited Number of Nodes ............................................................................................ 8
2.5. 2-D Rigid Body DOFs Subject to Applied Boundary Conditions .............................................................. 10
2.6. Rigid Sphere Translational DOFs + Rotational DOFs ............................................................................... 11
2.7. Rigid Body Translational DOFs Only ...................................................................................................... 11
2.8. MPC184 Universal Joint Geometry ........................................................................................................ 17
2.9. Stops Imposed on a Revolute Joint ........................................................................................................ 19
2.10. Stops Imposed on a Slot Joint .............................................................................................................. 20
2.11. Nonlinear Stiffness and Damping Behavior for Joints ........................................................................... 21
2.12. Coulomb's Law ................................................................................................................................... 22
2.13. Exponential Friction Law ..................................................................................................................... 22
2.14. Pinned Joint Geometry ....................................................................................................................... 26
2.15. Pinned Joint Mesh and Revolute Joint ................................................................................................. 27
2.16. Pinned Joint Contact Elements ............................................................................................................ 28
2.17. Pinned Joint Constraint Equations ....................................................................................................... 28
2.18. Rigid Constraint (KEYOPT(4) = 2) .......................................................................................................... 29
2.19. Flexible Constraint (KEYOPT(4) = 1) ...................................................................................................... 29
7.1. Overconstrained System: Standard 3-D Four-Bar Mechanism ................................................................. 60
7.2. Overconstraint Due to Redundant Rigid Components ............................................................................ 61
7.3. Overconstrained System: Cylindrical Tube Subjected to Bending at One End .......................................... 61
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. v
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
vi of ANSYS, Inc. and its subsidiaries and affiliates.
List of Tables
2.1. Rigid Body vs. Flexible Body Definition .................................................................................................. 12
2.2. Required Geometric Quantities ............................................................................................................. 23
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. vii
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
viii of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 1: Introduction to Multibody Simulation
Multibody simulation consists of analyzing the dynamic behavior of a system of interconnected bodies
composed of flexible (p. 5) and/or rigid (p. 6) components. The bodies may be constrained with respect
to each other via a kinematically admissible set of constraints modeled as joints. These systems can
represent an automobile, a space structure with antenna deployment capabilities, an aircraft as an as-
semblage of rigid and flexible parts, a robot with manipulator arms, and so on. In all such cases, the
components may undergo large rotation, large displacement, and finite strain effects.
This animated model of an aircraft landing gear is a typical example of a multibody simulation:
The following additional topics offer more information to help you understand multibody simulation
and how the Mechanical APDL program supports it:
1.1. Benefits of the Finite Element Method for Modeling Multibody Systems
1.2. Overview of the Multibody Analysis Process
1.3.The ANSYS-ADAMS Interface
1.4. Learning More About Multibody Dynamics
1.1. Benefits of the Finite Element Method for Modeling Multibody Sys-
tems
Multibody systems have conventionally been modeled as rigid body systems with superimposed elastic
effects of one or more components. These methods have been well documented in multibody dynamics
literature (p. 3). A major limitation of these methods is that nonlinear large-deformation, finite strain
effects, or nonlinear material cannot be incorporated completely into model.
The finite element (FE) method used in Mechanical APDL offers an attractive approach to modeling a
multibody system. While the Mechanical APDL multibody analysis method may require more computa-
tional resources and modeling time compared to standard analyses, it has the following advantages:
• The finite element mesh automatically represents the geometry while the large deformation/rotation effects
are built into the finite element formulation.
• Inertial effects are greatly simplified by the consistent mass formulation or even point mass representations.
• Interconnection of parts via joints (p. 13) is greatly simplified by considering the finite motions at the two
nodes forming the joint element.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 1
Introduction to Multibody Simulation
• The parameterization of the finite rotation has been well documented in the literature and can be easily
incorporated into the joint element formulations thereby enabling complete simulation of a multibody
system.
Mechanical APDL has an extensive library of elements available for modeling (p. 5) the flexible, rigid,
and joint components. You can model the material behavior of the flexible components using one of
several material models. Mechanical APDL also provides modal and transient dynamics capabilities to
analyze the spatial and temporal effects (p. 31) in a multibody simulation. Extensive postprocessing
capabilities are also available to interpret the analysis results (p. 37).
You can perform multibody simulation on a wide variety of mechanical systems. Typical applications
include automobiles and automobile components, aircraft assemblages, spacecraft applications, and
robotics.
Rigid bodies (p. 6) are modeled using MPC184 Rigid Link or Rigid
Beam elements, or by using the extensive contact capabilities
available in Mechanical APDL.
The flexible and/or rigid parts are connected using MPC184 joint
elements. For example, two parts may be simply connected such
that the displacements at the joining position are identical. In
other cases, the connection between two parts may involve a
more sophisticated joint such as the planar joint or universal joint.
In modeling these joints, suitable kinematic constraints (p. 31) are
imposed on the relative motion (displacement and rotation)
between the two nodes forming the joint.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
2 of ANSYS, Inc. and its subsidiaries and affiliates.
Learning More About Multibody Dynamics
A drawback of the ADAMS program is that all components are assumed to be rigid. In the ADAMS
program, tools to model component flexibility exist only for geometrically simple structures. To account
for the flexibility of a geometrically complex component, ADAMS relies on data transferred from finite-
element programs such as ANSYS. The ANSYS-ADAMS Interface is a tool provided by ANSYS, Inc. to
transfer data from the Mechanical APDL program to the ADAMS program.
For more information, see Appendix A: Rigid Body Dynamics and the ANSYS-Adams Interface in the
Substructuring Analysis Guide.
Current versions of Mechanical APDL support multibody analysis without the need for third-party tools.
Also, Mechanical APDL allows both rigid (p. 6) and flexible (p. 5) components.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 3
Introduction to Multibody Simulation
Geradin, Michel, and Alberto Cardona. Flexible Multibody Dynamics--A Finite Element Approach. New
York: Wiley, 2001.
Shabana, Ahmed A. Dynamics of Multibody Systems. 3rd ed. New York: Cambridge, 1998.
Clough, Ray W., and Joseph Penzien. Dynamics of Structures. Boston: McGraw-Hill, 1975.
Haug, Edward. Computer-Aided Kinematics and Dynamics of Mechanical Systems. Ed. Allyn & Bacon.
New Jersey: Prentice Hall, 1989.
Goldstein, Herbert, et al. Classical Mechanics. Boston: Addison-Wesley, 1950.
Kane, Thomas R., and David A. Levinson. Dynamics: Theory and Applications. Boston: McGraw-Hill,
1985.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
4 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 2: Modeling in a Multibody Simulation
A variety of issues can arise when modeling a multibody mechanism. The finite element modeling of
a multibody mechanism depends on the degree of complexity that you require. For example, it is often
possible to create a quick, initial approximation of the flexible (p. 5) and rigid (p. 6) parts of a
mechanism using standard beam elements and rigid beam/link elements. Alternatively, you can perform
detailed modeling of the flexible part using 3-D solid elements (or shell or solid-shell elements), and
the rigid part using the Mechanical APDL program's extensive contact capabilities.
Co
nn
ec
k
tin
an
gL
Cr
ink
Revolute Joint
{
I J
BE
AM
Fle
Be 4
18
18
am
xib 8
PC
le
Be
M
am Grounded
d
gi
Slot Joint
Ri
Grounded
Revolute Joint
• The rigid crank is connected to ground with a “grounded” MPC184 Revolute Joint element.
• The rigid crank and the connecting link are connected to each other by a MPC184 Revolute Joint element.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 5
Modeling in a Multibody Simulation
• The connecting link moves within a “grounded” MPC184 Slot Joint that approximates a slider block.
As a quick first attempt, you can model the flexible mechanism with some simple approximations to
the flexible and rigid parts. You can also model the connecting link in detail to study the deformation,
stresses, etc.
Mechanical APDL offers an extensive library of beam, shell, solid-shell, and solid elements for modeling
the flexible parts, and the extensive contact capability to model the rigid part and any other contact
conditions. Joint elements implemented via the Lagrange multiplier method offer the required kinematic
connectivity between any two parts or components.
To model mass and rotary inertia (p. 9), use the MASS21 element. The element is also appropriate for
use in a lumped approximation of rigid bodies.
Detailed information about element selection for flexible components is available in the Basic Analysis
Guide and the Element Reference.
In a finite-element model, certain relatively stiff parts can be represented by rigid bodies when stress
distributions and wave propagation in such parts are not critical. An advantage of using rigid bodies
rather than deformable finite elements is computational efficiency. Elements that belong to the rigid
bodies have no associated internal forces or stiffness. The motion of the rigid body is determined by a
maximum of six degrees of freedom (DOFs) at the pilot node.
For transient dynamic analyses, stiff bodies can excite high-frequency modes, resulting in a small time
increment in order to obtain a stable solution. Rigid bodies do not, however, excite any frequency
modes; therefore, using rigid bodies to represent stiff regions may allow a relatively large time increment.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
6 of ANSYS, Inc. and its subsidiaries and affiliates.
Modeling Rigid Bodies in a Multibody Analysis
The 3-D target element (TARGE170) and 2-D target element (TARGE169) are applied on the exterior
surface of the rigid body. To generate the target elements, issue an ESURF command.
The rigid body can also be a simple standalone body when the target elements do not overlap other
elements (that is, have no underlying elements), as shown:
You can generate target elements TARGE170 for a standalone 3-D rigid body (AMESH) or target elements
TARGE169 for a standalone 2-D rigid body (LMESH).
The most efficient rigid body should contain a limited number of nodes which are either connected to
other elements or subject to boundary conditions, as shown:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 7
Modeling in a Multibody Simulation
The rigid body shown above contains three nodes which connect five elements (two 3-D line segments,
one pilot node segment, one MASS21, and one MPC184-Revolute).
Target element POINT segments (TSHAP,POINT) can be defined and used to apply boundary conditions
(point loads, displacement constraints, etc.) on the rigid body surface where no predefined nodes exist.
2.2.1.2. Target Element Key Option Setting for Defining a Rigid Body
Each rigid body contains target elements defined by the same real constant ID. The target elements
can be defined via different element type IDs, however, you must set KEYOPT(2) = 1 on all of the target
elements. This KEYOPT setting causes Mechanical APDL to build internal multipoint constraints (MPC)
to enforce kinematics of the entire rigid body.
You can also combine different target segment types for each rigid body. However, you cannot mix 2-
D with 3-D target elements.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
8 of ANSYS, Inc. and its subsidiaries and affiliates.
Modeling Rigid Bodies in a Multibody Analysis
The pilot node, unlike the other segment types, is used to define the degrees of freedom for the entire
rigid body. This node can be any of the target element nodes, but it does not have to be. All possible
rigid motions of the rigid body will be a combination of a translation and a rotation around the pilot
node. The pilot node provides a convenient and powerful way to assign boundary conditions such as
rotations, translations, moments, temperature, voltage, and magnetic potential on the entire rigid body.
The pilot node can be connected to point mass, follower, and deformable elements. For a transient
analysis, you can simply locate the pilot node at the gravity center of the rigid body if the center of
mass is known.
Sometimes, the location of gravity center, the mass, and rotary inertia cannot be easily estimated. In
such cases, you can use the premeshed body to account for mass distribution for the rigid body (as
shown in Figure 2.2: Rigid Body Definition With Underlying Elements (p. 7)). The discretized elements
can be pure elastic solid, shell, or beam elements.
For each rigid body, you can perform the following steps:
4. Calculate inertia relief terms and print a summary of the mass properties (PSOLVE,ELPREP)
Based on the precalculated mass properties, you can easily define the point mass element. The node
is defined in the local coordinate system, as shown:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 9
Modeling in a Multibody Simulation
Set MASS21 KEYOPT(2) = 1 so that the point mass element coordinate system is initially parallel to the
nodal coordinate system and rotates with the nodal coordinate rotations during a large-deflection
analysis.
The DOFs of rigid body nodes are based on the DOFs of the connected elements and applied boundary
conditions (BCs). Rigid body nodes that connect to solid elements involve only the translational degrees
of freedom. Rigid body nodes that connect to shell, beam, follower, and joint elements also involve the
rotational DOFs.
For standalone rigid body nodes not connected to any other elements, the associated DOFs are subject
to applied boundary conditions, as shown:
Figure 2.5: 2-D Rigid Body DOFs Subject to Applied Boundary Conditions
The node has DOF UX if a constraint or a force is applied in the X direction. If there are no applied BCs,
the standalone rigid body nodes have no DOFs; in such a case, Mechanical APDL simply updates the
position of the nodes based on the kinematics of the rigid body.
The DOFs for a rigid body can also be controlled via KEYOPT(4) of the target element (TARGE169 or
TARGE170). The key option offers additional flexibility by fully or partially constraining the DOFs for the
rigid body.
Examples
In the following figure, a rigid sphere is defined by 8-node quadrilateral segments and a pilot node.
Two beam elements are connected to the rigid surface in the XY plane, as shown by the dotted lines.
The pilot node is located at the global Cartesian origin and is subjected to rotation ROTZ.
For the DOFs of the rigid body, selecting three rotational DOFs along with three translational DOFs rotates
the beams, as shown. Because the beams are fully connected to the rigid sphere, they rotate with the
sphere.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
10 of ANSYS, Inc. and its subsidiaries and affiliates.
Modeling Rigid Bodies in a Multibody Analysis
Selecting only the three translational DOFs for the rigid body, as shown in the following figure, does
not rotate the beams because they are connected only in their translational DOFs; therefore, the con-
nection acts as a hinge.
Determining the DOFs for each rigid body node is important because the internal multipoint constraints
are built solely on the resulting DOFs.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 11
Modeling in a Multibody Simulation
You can apply surface loads on surface effect elements SURF153 and SURF154 which fully or partially
override loads on the surface of the rigid body (p. 10).
Loads on a rigid body are assembled from contributions of all loads on nodes and elements connected
to the rigid body.
By selecting or deselecting target elements or the flexible finite elements, you can easily switch back
and forth between rigid body and flexible body definition.
The following table shows the general steps involved when defining a rigid body as compared to defining
a flexible body (p. 5):
2. Perform a partial element solution to obtain 2. Reselect the associated finite elements.
mass properties (p. 9).
3. Define the material properties for the flexible body.
3. Add a point mass element to the center of
rigid body. 4. Define a pilot node at one end of the joint. The pilot
node connects the joint to the rest of the body.
4. Add a target element whose node (pilot node)
shares the point mass node. 5. Select the nodes on the exterior surface of the body
that you want to connect to this pilot node.
5. Generate target elements on the exterior
surface of the pre-mesh body. 6. Create target elements on this surface.
6. Unselect the associated finite elements. For each body-joint connection, repeat steps 4
through 6. For more information, see Connecting
7. Connect joint elements to target nodes. Bodies to Joints (p. 26).
Caution
Redundant constraints are most likely to occur when two rigid bodies are connected to more
than one joint element.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
12 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
You cannot use the multipoint constraint (MPC) algorithm (KEYOPT(2)) and bonded or no-separation
contact behavior (KEYOPT(12)) to connect two rigid surfaces; doing so would cause the model to be
overconstrained, resulting in an abnormal termination of the analysis. You can simply replace the bonded
contact pair by adding an additional rigid body which connects two pilot nodes.
Mechanical APDL allows two rigid bodies that are connected or overlap each other through rigid body
nodes or the pilot node. To prevent overconstraints, the program merges two rigid bodies into one rigid
body internally and treats the second pilot node as a regular rigid body node.
MPC bonded contact between a flexible body and a rigid body is possible. The contact surface in an
MPC bonded contact pair, however, should always belong to the flexible body; otherwise, the MPC
bonded constraints and rigid body constraints are redundant.
An MPC184 joint element is defined by two nodes with six degrees of freedom at each node (for a total
of 12 DOFs). The relative motion between the two nodes is characterized by six relative degrees of
freedom. Depending on the application, you can configure different kinds of joint elements by imposing
appropriate kinematic constraints on any or some of these six relative degrees of freedom. For example,
to simulate a revolute joint, the three relative displacement degrees of freedom and two relative rota-
tional degrees of freedom are constrained, leaving only one relative degree of freedom available (the
rotation around the revolute axis). Similarly, constraining the three relative displacement degrees of
freedom and one relative rotational degree of freedom can simulate a universal joint. Two rotational
degrees of freedom are “unconstrained” in this joint.
The kinematic constraints in the joint elements are imposed using the Lagrange multiplier method.
Because the Lagrange multiplier method is used to impose the constraints, the constraint forces are
available for output purposes.
The following topics about using joint elements in a multibody analysis are available:
2.3.1. Joint Element Types
2.3.2. Material Behavior of Joint Elements
2.3.3. Reference Lengths and Angles for Joint Elements
2.3.4. Boundary Conditions for Joint Elements
2.3.5. Connecting Bodies to Joints
The following table lists the different types of joint elements and the required key option settings. The
relevant element section in the Element Reference is also indicated.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 13
Modeling in a Multibody Simulation
Joint Element Type KEYOPT(1) KEYOPT(4) MPC184 Element Constraints (p. 62)
Revolute (p. 15) 6 ---
Revolute Joint 5
Z-axis revolute 6 1
Universal (p. 15) 7 --- Universal Joint 4
Slot (p. 15) 8 --- Slot Joint 2
Point-in-plane 9 --- Point-in-Plane Joint 1
Translational (p. 15) 10 --- Translational Joint 5
Cylindrical (p. 15) 11 ---
Cylindrical Joint 4
Z-axis cylindrical 11 1
Spherical (p. 16) 5 --- Spherical Joint 3
Planar (p. 16) 12 ---
Planar Joint 3
Z-axis planar 12 1
Weld 13 --- Weld Joint 6
Orient 14 --- Orient Joint 3
General 16 --- General Joint Depends
on
number
of fixed
relative
DOFs
Minimum
constraints
= 0 (No
DOF is
fixed)
Maximum
constraints
= 6 (All
DOFs
are
fixed)
Screw 17 --- Screw Joint 5
Relative
axial
motion
and
rotational
motion
are
linked
via the
pitch of
the
screw
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
14 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
Revolute Joint Constrained degrees of freedom: UX, UY, UZ, ROTX, ROTY
Translational Joint Constrained degrees of freedom: UY, UZ, ROTX, ROTY, ROTZ
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 15
Modeling in a Multibody Simulation
Example
If the first node of the joint element is a grounded node, then the element definition is: E,,J or
EN,ElementNumber,,J
Similarly, if the second node is the grounded node, then the element definition is: E,I, or EN,Element-
Number,I
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
16 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
Example
The local coordinate systems and their required orientation vary from one joint element to another.
Input data requirements for each joint element differ. Typically, the local coordinate system is always
defined at the first node of a joint element.
The local coordinate system at the second node may be optional. If it is not specified, then the local
coordinate system at the first node is usually assumed.
The rotational components of the relative motion between the two nodes of the joint elements are
quantified in terms of Bryant (or Cardan) angles that are evaluated based on these coordinate systems.
Example
The following figure illustrates the specification of the local coordinate system for a universal joint ele-
ment:
J
e3
eI
2
eJ
2
eI e J1
1
eI
3 J
e3
eI
2
I
Z J
e J2
Y
X I, J - Coincident nodes
eI e J1
1
eI J
I 3 e3
J
eI
2
eJ
2
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 17
Modeling in a Multibody Simulation
LOCAL,11,0
...
LOCAL,12,0
...
SECJOINT,LSYS,11,12
Caution
Use joint stops sparingly. The program treats a stop constraint internally as a "must be im-
posed" or "hard" constraint, and no contact logic is used. As a result, during the given iteration
of a substep, the stop constraints activate immediately if the program detects a violation of
a stop limit.
Depending upon the nature of the problem, the stop constraint implementation may cause
the solution to trend towards an equilibriated state that may not be readily apparent to you.
In addition, do not use stops to simulate zero-displacement boundary conditions. You should
also avoid specifying stops on multiple joints.
Finally, do not use joint stops as a substitute for contact modeling. Whenever possible, use
node-to-node or node-to-surface contact modeling to simulate limit conditions.
In a transient dynamic analysis, if relative displacement-based (or rotation-based) stop constraints are
used, then the relative velocities and relative accelerations become inconsistent (oscillatory velocity
and/or accelerations are observed in many cases), implying that the energy and momentum due to the
impact-like nature of the stops is not conserved. These inconsistencies are reasonably suppressed by
imposing a numerical damping. However, numerical damping does not work appropriately in some
cases. Thus, for the transient dynamic case, an energy-momentum conservation scheme is adopted. By
this method, the user specified relative DOF stop values are taken into account, and constraints based
on the relative velocity are imposed in such a way that the overall energy and momentum balance is
achieved in a finite element sense.
Irrespective of the integration scheme specified for the transient dynamic analysis, the Newmark
method is used for the joint element when stops are specified.
The energy-momentum conservation scheme for stops is implemented for all joints except the screw
joint. In the case of the screw joint, the stops are imposed based on the relative displacements (or ro-
tations).
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
18 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
You can impose stops or limits on the available components of relative motion between the two nodes
of a joint element. The stops or limits essentially constrain the values of the free DOFs within a certain
range. To specify minimum and maximum values, issue the SECSTOP command.
The following figure shows how stops can be imposed on a revolute joint such that the motion is
constrained. The axis of the revolute is assumed to be perpendicular to the plane of paper and is along
the e3 direction.
The local coordinate system specified at node I is assumed to be fixed in its initial configuration. However,
the local coordinate system specified at node J evolves with the rotation of that node. The relative
angle of rotation is given by:
Let the link with node J rotate with respect to the link with node I. This characteristic implies that the
local coordinate system at node J rotates with respect to the local coordinate system at node I.
For the configuration shown, the initial relative angle of rotation is zero degrees. A counterclockwise
motion results in positive angles of rotation. Clockwise motion results in negative angles of rotation.
If stops limit the movement of the link with node J (as shown), the stop conditions are specified as
follows:
SECSTOP,6,PHImin ,PHImax
The next figure shows how stops can be imposed in a slot joint which involves displacements in the
local eI 1 axis of node I. The relative distance between node J and node I is given by:
where xI and xJ are the position vectors of nodes I and J. The initial distance between the nodes I and
J is l 0 and is a positive value.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 19
Modeling in a Multibody Simulation
Referring to Figure 2.9: Stops Imposed on a Revolute Joint (p. 19), the locks for a revolute joint are
specified as SECLOCK,6, Phi_Min,Phi_Max
Referring to Figure 2.10: Stops Imposed on a Slot Joint (p. 20), the locks for the slot joint are specified
as SECLOCK,1,l_Min,l_Max
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
20 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
The nonlinear stiffness and damping behavior is specified using the TB,JOIN command with an appro-
priate TBOPT label. In the case of nonlinear stiffness, relative displacement (rotation) versus force
(moment) values are specified using the TBPT command. For nonlinear damping behavior, velocity
versus force behavior is specified using the TBPT command. (See Figure 2.11: Nonlinear Stiffness and
Damping Behavior for Joints (p. 21) for a representation of the nonlinear stiffness or damping curve.)
In either case, these values may be temperature dependent; use the TBTEMP command to define the
temperature for the data table.
You can specify the linear or nonlinear stiffness or damping behavior independently for each component
of relative motion. However, if you specify linear stiffness for an unrestrained component of relative
motion, you cannot specify nonlinear stiffness behavior on the same component of relative motion.
The damping behavior is similarly restricted. If a joint element has more than one free or unrestrained
component of relative motion--for example, the universal joint has two free components of relative
motion--then you can independently specify the stiffness or damping behavior as linear or nonlinear
for each of the unrestricted components of relative motion.
The laws governing the frictional behavior of the joint are described below.
Coulomb’s Law
The classical Coulomb friction model is implemented for joints using a penalty formulation. The Coulomb
friction model for joints is defined as:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 21
Modeling in a Multibody Simulation
Where, F s is the equivalent tangential force (or moment), F n is the normal force (or moment) in the
joint, and μ is the current value of the coefficient of friction. The calculation of the normal force depends
on the joint under consideration.
If the equivalent tangential force F s is less than F lim , the state is known as the sticking state. If F s
exceeds F lim , sliding occurs and the state is known as the sliding state. The sticking/sliding calculations
determine when a point transitions from sticking to sliding or vice versa.
Fs
Sliding
Flim
Fn
The exponential friction law is used to smooth the transition between the static coefficient of friction
and the dynamic coefficient of friction according to the formula (Benson and Hallquist):
where:
c = decay coefficient
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
22 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
If appropriate geometric quantities are not specified, then the corresponding normal force contributions
will not be considered. The following section explains the normal force calculations and the geometric
quantities required.
Revolute Joint
In order to compute the normal moment in a revolute joint, the revolute joint is visualized as a cylinder-
pin assembly (for example, a door hinge consisting of a pin with a head inserted into a cylinder).
The following geometric quantities are required in the calculations below. Note that the specification
of these quantities is optional. If some of these geometric quantities are not specified, then the corres-
ponding contribution to the normal moment calculations is ignored.
• L eff = The effective length is the length over which the cylinder and pin are in contact with each other
The contributions to the normal moment in an x-axis revolute joint are as follows:
• An axial moment due to the axial component of the constraint Lagrange Multiplier force (λ1 ).
This force acts in such a way as to push the cylinder against the pin head, thereby causing a frictional
moment to develop.
where,
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 23
Modeling in a Multibody Simulation
• A tangential moment due to the constraint Lagrange Multiplier forces, λ2 and λ3:
• A bending moment that is generated as a consequence of the constraint Lagrange Multiplier moments (λ5
and λ6):
Additionally, if interference fit moment (M interference ) is defined, the normal moment for frictional cal-
culations is given by:
A similar calculation is carried out for the z-axis revolute joint by choosing the appropriate constraint
Lagrange multiplier forces in the above equations.
Slot Joint
The two displacement constraint Lagrange Multiplier forces (λ2 and λ3) in the slot joint contribute to
a tangential force as follows:
Additionally, if interference fit force (F interference ) is defined, the normal force for frictional calculations
is given by:
Translational Joint
• L eff = Effective length. The effective length is the length over which the two parts of the translation joint
overlap. It is assumed that the change in this length is small.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
24 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
• R eff = Effective radius. To simplify calculations, an effective radius is used in torsional moment calculations,
even though the cross section in a translational joint is rectangular. The effective radius is used in computing
the force that arises due to the torsional moment.
• An effective radial force due to the constraint forces (λ2 and λ3):
Additionally, if interference fit force (F interference ) is defined, the normal force for frictional calculations
is given by:
If you do not define reference lengths and angles, Mechanical APDL calculates the values from the initial
configuration of the joints. The program uses the reference lengths and angles in the stiffness and
frictional behavior calculations.
To apply concentrated forces on the available components of relative motion of the joint element, issue
the FJ command. You can list the imposed values via the FJLIST command. To delete the values, issue
the FJDELE command.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 25
Modeling in a Multibody Simulation
Figure 2.14: Pinned Joint Geometry (p. 26) shows a 3-D model of a pinned joint where the geometry
of the joint (the pin) is explicitly modeled. To perform a multibody analysis, the pin geometry is ignored
and the behavior replaced by the appropriate MPC184 joint element (p. 13).
Figure 2.15: Pinned Joint Mesh and Revolute Joint (p. 27) shows the meshed model including the revolute
joint. To connect the bodies to the joint, you must use either elements (such as beams) or constraint
equations. The easiest way to do so is to use contact elements to create surface-based constraints
(multipoint constraints, or MPCs), as follows:
1. Define a pilot node at one end of the joint. The pilot node connects the joint to the rest of the body.
2. Select the nodes on the surface of the body that you want to connect to this pilot node.
3. Create contact surface elements on this surface. By sharing the same real constant number (REAL,N ), MPCs
between the surface nodes and the pilot node are generated during the solution.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
26 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
Figure 2.16: Pinned Joint Contact Elements (p. 28) shows the contact elements and Figure 2.17: Pinned
Joint Constraint Equations (p. 28) shows the MPCs (constraint equations) created during the solution
for the lower body.
Create the pilot node using the TARGE170 element--setting KEYOPT(2) = 1 so as not to allow the program
to constrain any DOFs--and issuing the TSHAP,PILO command.
If you mesh the body with elements having no midside nodes (such as SOLID185), use CONTA173 as
the element type for the surface mesh. For elements with midside nodes (such as SOLID186 or SOLID187),
use CONTA174. Set the following element key options to create the necessary constraints:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 27
Modeling in a Multibody Simulation
Instead of the rigid option, you can also choose a flexible (force-distributed or RBE3-type) constraint
option by setting KEYOPT(4) = 1. The following figures illustrate the difference in behaviors:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
28 of ANSYS, Inc. and its subsidiaries and affiliates.
Connecting Multibody Components with Joint Elements
Contact elements
Contact elements
Additional Information
For more information about using contact elements to generate constraints, see Surface-Based Constraints
in the Contact Technology Guide.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 29
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
30 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 3: Performing a Multibody Analysis
A multibody refers to a structural system consisting of flexible (p. 5) and rigid (p. 6) components.
The following structural analysis types are available for multibody analysis: static, modal, harmonic,
transient dynamic, spectrum, and buckling. For more information about each supported structural
analysis type, see the Structural Analysis Guide.
The following topics present information necessary for performing a successful multibody analysis:
3.1. Kinematic Constraints
3.2. Convergence Criteria
3.3. Initial Conditions
3.4. Damping
3.5.Time-Step Settings
3.6. Solver Options
Providing sufficient kinematic constraints for a finite element model would lead to a full rank system
of equations which would give a unique solution. Lack of sufficient kinematic constraints would make
the system unstable. A finite element solution for such a system would fail to converge.
If more than sufficient kinematic constraints are specified for the structural system, the system may remain
stable or become unstable. If the extra constraints conflict with the basic constraints necessary to keep
the system in stable equilibrium, the system becomes unstable and the finite element solution fails
with convergence problems. If the extra constraints do not conflict with the basic constraints, the system
is consistently overconstrained and the extra constraints become redundant constraints. The system re-
mains stable; however, there is no unique solution. Depending on how the equations for the finite
element model are solved, the solution may or may not converge.
To ensure convergence of the finite element solution, the system must not be underconstrained or
overconstrained. Checking for either lack of sufficient constraints or overconstraints can be difficult for
complex systems. ANSYS, Inc. recommends performing a modal analysis on the system. If the modal
analysis yields more zero eigenvalues than the rigid body modes of the system, the system lacks sufficient
constraints; if there are fewer eigenvalues than rigid body modes, the system is overconstrained. A
closer look at the unwanted eigenmodes can point to the missing or extra constraints.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 31
Performing a Multibody Analysis
Because all time-integration schemes (such as the Newmark method and the HHT method) rely on the
history of displacements, velocities and accelerations, it is important to define consistent initial conditions.
By default, a zero value is assumed for initial displacements, velocities, and accelerations at DOFs that
are not otherwise specified (via the IC command).
Inconsistencies in initial conditions introduce errors into the time-integration scheme and lead to excit-
ation of undesired (spurious) modes. Accumulation of these errors over several time increments adversely
affects the solution and very often causes the time-integration scheme to fail. Applying numerical
damping (p. 35) or other forms of damping can suppress the growth of these errors. However, such
additions also affect the solution, especially, when long term transient behavior is being studied in the
analysis.
It is not always possible, however, to have complete information about the initial state of a system being
modeled for transient analysis. In such situations, it is helpful to run a dummy load step before the ac-
tual transient analysis of interest to bring the system into a consistent initial state. The purpose of such
a load step is to eliminate the error introduced by inconsistent initial conditions.
Example
Consider a rigid beam of length l rotating in the x-y plane about a pinned end at a constant angular
velocity ω. The free end of the beam has a tangential velocity of ωl and a centripetal acceleration of
ω2l. The beam is assumed to have all of its mass concentrated at the free end. To perform the analysis,
model the rigid beam using the MPC184 element with Lagrange multipliers to enforce the rigid beam
constraints. With one end of the rigid beam pinned, apply initial velocity normal to the beam axis at
the free end. To introduce centripetal acceleration, use acceleration loading as illustrated in the following
input file:
/prep7
et,1,mass21
keyopt,1,3,2 !3d mass without rotary inertia
et,2,mpc184
keyopt,2,1,1 !rigid beam
keyopt,2,2,1 !lagrange multiplier
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
32 of ANSYS, Inc. and its subsidiaries and affiliates.
Initial Conditions
type,1
real,1
m = 1.0
r,1,m
en,1,2 !3d mass at free end (node 2)
type,2
real,2
en,2,1,2 !rigid beam
finish
/solu
d,1,all
ddel,1,rotz
d,2,uz
d,2,rotx
d,2,roty
time,6.0
acel,0.0,0.0,0.0 !remove centripetal acceleration
kbc,1
midtol,on,1e2 !automatic time stepping with MIDTOL
nsub,600,1e7,400
trnopt,full, , , , ,HHT
tintp,0.05 !small numerical damping for HHT
outres,all,all
solve
finish
/post26
/xrange,0.,6.0
nsol,2,2,u,x,ux !x displacement for node 2
nsol,3,2,u,y,uy !y displacement for node 2
nsol,4,2,v,x,vx !x velocity for node 2
nsol,5,2,v,y,vy !y velocity for node 2
nsol,6,2,a,x,ax !x acceleration for node 2
nsol,7,2,a,y,ay !y acceleration for node 2
/axlab,x,Time T
/axlab,y,D/V/A
/gropt,divx,10
/gropt,divy,10
/gthk,curve,2
/title,Transient analysis of a rigid 3D beam rotating about a fixed node
plvar,ux,uy,vx,vy,ax,ay
finish
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 33
Performing a Multibody Analysis
Some potential drawbacks exist in cases where high frequency content of flexible multibody systems
is important for analysis. Applying high numerical damping in the dummy analysis can affect the desired
high-frequency response. ANSYS, Inc. recommends using the HHT method for this technique because
the integration scheme shows good dissipation properties with numerical damping.
Example
Consider a rigid-flexible double pendulum made up of a rigid and a flexible beam. One end of the rigid
beam is pinned and the other end is hinged to the flexible beam. The other end of the flexible beam
is free. The rigid beam is assumed to have all of its mass concentrated at the end that is hinged to the
flexible beam. The system is given an initial velocity tangential to the flexible beam axis at its free end,
as shown in the following input file:
/prep7
et,1,mass21
keyopt,1,3,2 !3d mass without rotary inertia
et,2,mpc184
keyopt,2,1,1 !rigid beam
keyopt,2,2,1 !lagrange multiplier
et,3,mpc184
keyopt,3,1,6 !revolute joint between rigid and flexible beam
et,4,beam188 !flexible beam
type,1
real,1
m = 390
r,1,m
en,1,2 !3d mass at the end of rigid beam
type,2
real,2
en,2,1,2 !rigid beam
sectype,4,beam,csolid
secdata,1,0.1784124116 !c-s area is 0.1
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
34 of ANSYS, Inc. and its subsidiaries and affiliates.
Damping
type,4
real,4
secnum,4
mat,1
en,4,3,4 !flexible beam elements
en,5,4,5
en,6,5,6
en,7,6,7
d,1,all
ddel,1,rotz
finish
/solu
time,6.0
midtol,on,10 !automatic time stepping with MIDTOL
nsub,100,1e6,100
trnopt,full, , , , ,HHT
tintp,0.05 !small numerical damping for HHT
outres,all,all
solve
finish
/post26
nsol,2,7,u,x,ux !x displacement for node 7
nsol,3,7,u,y,uy !y displacement for node 7
nsol,4,2,u,x,ux1 !x displacement for node 2
nsol,5,2,u,y,uy1 !y displacement for node 2
nsol,4,3,v,x,vx !x velocity for node 7
nsol,5,3,v,y,vy !y velocity for node 7
nsol,6,7,a,x,ax !x acceleration for node 7
nsol,7,7,a,y,ay !y acceleration for node 7
/axlab,x,Time T
/axlab,y,D/V/A
/gropt,divx,10
/gropt,divy,10
/gthk,curve,2
/title,Transient analysis of a rigid-flexible double pendulum
plvar,ux,uy,ux1,uy1,vx,vy,ax,ay
finish
3.4. Damping
You can specify two types of damping:
3.4.1. Numerical Damping
3.4.2. Structural Damping
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 35
Performing a Multibody Analysis
Numerical damping stabilizes the numerical integration scheme by damping out the unwanted high
frequency modes. For the Newmark method, numerical damping also affects the lower modes and reduces
the accuracy of integration scheme from second order to first order. For the HHT method, numerical
damping affects only the higher modes and always maintains second-order accuracy.
Mechanical APDL uses a default value (TINTP,GAMMA) of 0.005. The value that you select should be
based on the problem at hand. A sensible value to try initially is 0.1. Use the lowest possible value that
damps out nonphysical response without significantly affecting the final solution. Problems involving
rigid body translational motion, other forms of damping, or dissipative mechanisms like plasticity or
friction typically require smaller values for numerical damping. Larger numerical damping values are
usually necessary for problems involving rigid body rotational motion, elastic collisions (dynamic con-
tact/impact), and large deformations with frequent changes in substep size.
For faster convergence in a full transient dynamic analysis where mass elements such as MASS21 are
used, issue the NROPT,UNSYM command. The command activates the Newton-Raphson option for
solving the nonlinear equations in the analysis, necessary due to the nonsymmetric stiffness contribution
resulting from gyroscopic effects.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
36 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 4: Reviewing Multibody Analysis Results
Results from a flexible multibody analysis consist mainly of displacements, velocities, accelerations,
stresses, strains, and reaction forces in structural components. Constraint forces, current relative positions,
relative velocities, and relative accelerations in joint elements are also available.
Results are available for viewing in POST1, the general postprocessor (/POST1), or in POST26, the time-
history postprocessor (/POST26).
For a description of the available output components, see the Output Data sections of the element
descriptions for any of the elements that model the flexible components (p. 5), rigid components (p. 6),
and joint elements (p. 13).
The following topics concerning how to review flexible multibody analysis results are available:
4.1. Reviewing Results in POST1
4.2. Reviewing Results in POST26
4.3. Output of Joint Element Quantities
4.4. Energy Output
• The database must contain the same model for which the solution was calculated.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 37
Reviewing Multibody Analysis Results
Many other postprocessing functions are available in POST1. For more information, see The General
Postprocessor (POST1) in the Basic Analysis Guide.
A typical POST26 postprocessing sequence for a flexible multibody analysis is similar to the sequence
for a typical nonlinear analysis, as follows:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
38 of ANSYS, Inc. and its subsidiaries and affiliates.
Output of Joint Element Quantities
Many other postprocessing functions are available in POST26. For more information, see The Time-History
Postprocessor (POST26) in the Basic Analysis Guide.
• Stop status
• Lock status
• Relative position
• Relative velocities
• Relative accelerations
• The components of the bases vectors at the two nodes in the deformed configuration.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 39
Reviewing Multibody Analysis Results
The bases vectors are specified as the local coordinate systems via the SECJOINT command and
evolve with the rotation of the underlying nodes.
• The constraint forces and moments in the evolved basis at the first node of the joint element.
The ANSYS Workbench Products generally use NMISC output for postprocessing.
See the MPC184 element documentation and the individual joint element descriptions for details about
the SMISC component specification and the use of the ETABLE command.
In POST1 (p. 37), you can print joint element output (such as relative reaction forces, relative displace-
ments, relative rotations, etc.) at the free or unconstrained relative degree of freedom via the PRJSOL
command. To obtain the nodal forces at the joint element nodes, issue the PRESOL,FORC command.
In POST26 (p. 38), you can use the JSOL command to specify result items (such as relative displacements,
velocities, accelerations, etc.) that must be stored for a joint element. Then, you can plot or print the
stored items via the PLVAR or PRVAR command, respectively.
In POST26 (p. 38), you can use the ENERSOL command to store a specific energy item. Then, you can
graph or list the specific energy item in the output file via the PLVAR or PRVAR command, respectively.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
40 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 5: Using Component Mode Synthesis Superelements in a
Multibody Analysis
Obtaining the flexible response of a body or bodies to a dynamic motion event typically involves solving
hundreds or thousands of time points. If a flexible body has many degrees of freedom (DOFs), a multibody
analysis can be time-consuming. To minimize the necessary computing resources, you can use component
mode synthesis (CMS) superelements (substructures) to replace the many thousands of DOFs of the
flexible body with tens of DOFs that represent the dynamic response, thereby significantly reducing
the required multibody analysis run time.
The following topics describe the approach required to perform a substructure-based multibody analysis,
including recovering the time-dependent flexible response:
5.1. CMS Superelements in a Multibody Analysis
5.2. Flexible Body Types
5.3. Substructuring Overview
5.4. Master Degrees of Freedom in a Substructured Multibody Simulation
5.5. Steps for Performing a Substructured Multibody Simulation
For an example of how to set up and use a substructuring in a multibody analysis, see Example Multibody
Analysis: Crank Slot Mechanism (p. 51).
• Nonlinear elements within the body (such as gasket or contact elements) are treated as linear and in their
initial state.
• The body may consist only of 3-D structural elements. (You can use 2-D elements with care provided that
you follow the guidelines given later, particularly with respect to the number of DOFs at the master DOFs.)
The body may undergo large rotations, but the strains and relative rotations within the body are pre-
sumed to be small.
• Bodies that are excited by the motion of other bodies (rigid or flexible) but do not themselves undergo
large motions
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 41
Using Component Mode Synthesis Superelements in a Multibody Analysis
An engine block is an example of this type, where the block is excited dynamically from the crankshaft,
pistons, and other moving parts attached or linked to the block. This case is a straightforward applic-
ation of traditional superelements.
A piston rod is an example of a body undergoing large motions; this type also uses superelements
but with the additional capability that the superelement can undergo large motions, and large rotations
in particular. A large-rotation superelement involves additional considerations (p. 42).
• A generation pass, where the group of elements are condensed down to generate the superelement.
• A use pass, where the superelement is used in the analysis. In our case, in the multibody analysis.
• An expansion pass, where the results of the superelement in the use pass are expanded to the original group
of elements so that their displacements, forces, strains, and stresses are recovered.
In the use pass, Mechanical APDL allows the superelement to rotate with arbitrarily large rotations.
In the generation pass, you define master degrees of freedom (p. 42) (MDOFs). The MDOFs are the
DOFs that the superelement uses to interface with, or connect to, the other bodies or joints.
Because the flexible body analysis occurs within a dynamic analysis, you must include the dynamic
(mass) effects. Use component mode synthesis (CMS) to augment the superelement static stiffness with
mode shapes that characterize the dynamic behavior, much as you would when performing a mode-
superposition transient dynamic analysis.
CMS is a form of substructure analysis allowing you to derive the dynamic behavior of the entire assembly
from its constituent components. For more information, see Component Mode Synthesis in the Substruc-
turing Analysis Guide.
If the connection occurs at a joint at the center of a hole or slot, you must place a master node there.
For more information, see Connecting Bodies to Joints (p. 26).
Nonrotating Bodies
For nonrotating bodies, master nodes are located at the points where the superelement connects with
the other bodies and are typically located at the centers of bolts or other fasteners and bearings. Try
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
42 of ANSYS, Inc. and its subsidiaries and affiliates.
Steps for Performing a Substructured Multibody Simulation
to minimize the number of master nodes. Where appropriate, use the techniques presented in Connecting
Bodies to Joints (p. 26) to create a single master node that connects to a number of nodes.
Rotating Bodies
For rotating bodies, the idea is to create a beam-like superelement, ideally with two master nodes (but
never less than two). You can use more than two master nodes (for example, when modeling a lever
or rocker plate), but Mechanical APDL assumes that the rotation of the superelement is the average of
the rotations of all master nodes.
All master nodes of a rotating body must have six active structural DOFs: UX, UY, UZ, ROTX, ROTY, and
ROTZ. If the master node does not have six DOFs--for example, if it is the node of a 3-D solid element-
-create a six-DOF node at that location and tie it to the rest of the body appropriately. You can use
either of the following techniques, both of which essentially place a six-DOF node connected to a patch
of elements superimposed on the existing solid elements.
• MPC Contact -- Create a pilot node and link it to bonded contact elements overlaid on the patch. For more
information, see Connecting Bodies to Joints (p. 26).
• Beams -- Overlay beam elements or MPC184 Rigid Beam elements in a spiderweb fashion. The beams should
have high stiffness and no mass.
You can also define MDOFs where loads are to be applied as well as at any points where velocities or
accelerations are of interest.
Using substructures to represent some or all of the flexible bodies in a completely defined multibody
model requires the following steps:
5.5.1. Step 1: Prepare the Full Model for a Substructured Multibody Analysis
5.5.2. Step 2: Create the Substructures (Generation Pass)
5.5.3. Step 3: Build the CMS-based Model (Use Pass)
5.5.4. Step 4: Run the Multibody Analysis
5.5.5. Step 5: Expand all Solutions (Expansion Pass)
5.5.6. Step 6: Create the Merged Results File
5.5.7. Step 7: Postprocess the Results
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 43
Using Component Mode Synthesis Superelements in a Multibody Analysis
Before proceeding, prepare the full multibody model (as described in Steps 1 through 4 in Overview
of the Multibody Analysis Process (p. 2)). Verify that the bodies are connected to the joints as described
in Connecting Bodies to Joints (p. 26).
The multiple passes used in substructuring require that the files created and used in the process are
handled appropriately. To aid in file management when performing a substructured multibody simulation,
use the /FILNAME command to modify the current jobname as needed.
5.5.1. Step 1: Prepare the Full Model for a Substructured Multibody Analysis
Prepare the full model for a substructured multibody analysis, as follows:
You must decide how many modes to include in the CMS substructure. The number you determine
depends on several factors including:
• Whether displacements are of primary interest, or whether stresses/strains are of primary interest. (The latter
require more modes to accurately capture their response.)
For most analyses, and particularly for rotating bodies, the fixed-interface method (CMSOPT,FIX) is
sufficient. For analyses where higher frequencies are of interest (foe example, those involving acoustics
or high-speed equipment), the residual-flexible free-interface method (CMSOPT,RFFB) provides more
accuracy. For more information, see Supported CMS Methods in the Substructuring Analysis Guide.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
44 of ANSYS, Inc. and its subsidiaries and affiliates.
Steps for Performing a Substructured Multibody Simulation
For nonrotating bodies, you can apply constraints (D) in the generation pass to the degrees of freedom
(DOFs), but not the master degree of freedom (MDOF). Set KEYOPT(4) = 1 for these superelements in
the use pass; otherwise, your analysis will have convergence problems. For rotating bodies, do not apply
constraints in the generation pass because the superelement must have six rigid body modes; you can,
however, apply constraints to its MDOF in the use pass.
Loading Considerations
• The loads rotate with the rotating substructure by default. This behavior is valid for most load types (especially
pressure loads). In the use pass, however, you can specify that the load vector not rotate with the substructure;
disabling load rotation is useful in some cases, such as those involving nodal forces where you want to
maintain their original direction.
• When to apply gravity and other acceleration loads (such as those applied via ACEL and OMEGA commands)
depends on whether the body is rotating or not. For a rotating body, apply the loads in the use pass. For a
nonrotating body, you can apply the loads in this step and use it in the use pass; however, be careful not
to specify it twice (for example, by issuing an ACEL command in the use pass). Issue the CMACEL command
to apply the acceleration to the nonsubstructured elements only.
• By applying a unit load in this step, you can easily scale it in the use pass and make use of tabular loads to
apply a complex load-versus-time history in a single load step. ANSYS, Inc. recommends this approach as it
allows for straightforward creation of the full model results file.
Follow these steps to create the superelements for a substructured multibody analysis:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 45
Using Component Mode Synthesis Superelements in a Multibody Analysis
Repeat the steps above for each flexible body you wish to replace with CMS substructures. Use unique
jobnames and substructure names for each flexible body.
If you are using the residual-flexible free interface method, use CMSOPT,RFFB,NMODE (rather than CM-
SOPT,FIX,NMODE) in Step 2.5. You must also define pseudo-constraints (D,,,SUPPORT).
For further information, see The CMS Generation Pass in the Substructuring Analysis Guide.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
46 of ANSYS, Inc. and its subsidiaries and affiliates.
Steps for Performing a Substructured Multibody Simulation
Caution
Be careful not to select all elements (for example, via an ALLSEL command) before initiating
the solution (SOLVE) in the next step. If you do so, Mechanical APDL solves for both sets of
elements.
To dampen out excessive solution noise, particularly in the velocities and accelerations, you typically
use numerical damping. For more information, see Damping (p. 35).
In Step 4.3, use tabular loads to specify complex load-versus-time histories. By default, loads are simply
ramped (or step-applied [KBC]) over the time interval from one load step to the next. Tabular loads,
however, allow a general load curve. To use multiple load steps to define the loading, repeat Steps 4.3
and 4.4 for each load configuration.
For more information about setting up and performing a multibody analysis, see Performing a Multibody
Analysis (p. 31).
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 47
Using Component Mode Synthesis Superelements in a Multibody Analysis
Repeat all steps for each substructured body (including clearing the database [/CLEAR]).
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
48 of ANSYS, Inc. and its subsidiaries and affiliates.
Steps for Performing a Substructured Multibody Simulation
• NSUBSTEPS is the total number of substeps (time points) in the results files.
• In the example commands, the jobname from the use pass (Step 3 (p. 46)) is USE; therefore, its results file
is named USE.RST. Likewise, the jobname from the expansion pass (Step 5 (p. 47)) is BODY1; therefore, its
results file is named BODY1.RST. Adjust the command arguments accordingly to accommodate your own
jobnames.
• As presented here, the analysis in the use pass is performed in one load step with NSUBSTEPS substeps. If
such is not the case in your analysis, modify the *DO loop to use the appropriate SET command.
• The expansion pass results files always have only one load step with all time points contained as NSUBSTEPS
substeps, irrespective of the use pass load stepping and substepping.
Use the POST1 postprocessor (/POST1) to review the results over the entire model. Use the POST26
postprocessor (/POST26) to obtain time-history listings and plots. For more information, see Reviewing
Multibody Analysis Results (p. 37) for specific multibody postprocessing.
Nodal velocity and acceleration nodal results are not available for the substructure interior nodes (non-
MDOFs).
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 49
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
50 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 6: Example Multibody Analysis: Crank Slot Mechanism
The example crank slot analysis in this section introduces you to the Mechanical APDL program's
multibody analysis capabilities. To facilitate modeling and simulation in a multibody analysis, ANSYS,
Inc. suggests using the ANSYS Workbench product along with Mechanical APDL to develop your ana-
lysis. The input files (p. 58) used to run the crank slot analysis in Mechanical APDL were generated by
ANSYS Workbench.
The following topics are available for this example multibody analysis of a crank slot mechanism:
6.1. Problem Description
6.2. Problem Specifications
6.3. Defining Joints
6.4. Performing the Rigid Body Analysis
6.5. Performing the Flexible Body Analysis
6.6. Using Component Mode Synthesis in the Multibody Analysis
6.7. Using Joint Probes
6.8. Comparing Processing Times
6.9. Input Files Used in This Analysis
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 51
Example Multibody Analysis: Crank Slot Mechanism
The following figure shows the parts of the model, with the joints listed to the right:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
52 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing the Rigid Body Analysis
All joints are available via the MPC184 element's KEYOPT(1) setting and, in some cases, the KEYOPT(4)
setting. For more information, see Connecting Multibody Components with Joint Elements (p. 13).
• MPC184 elements for the joints connected to each other via rigid body nodes.
For more information, see Modeling Rigid Bodies in a Multibody Analysis (p. 6). The input file (p. 58)
CrankSlot_Rigid.inp (available on your product distribution media) is used to perform the rigid
body portion of the analysis.
The following figures show the finite element (FE) representation of the model and the time-history
plot of the total displacement of the rigid Rod2 part:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 53
Example Multibody Analysis: Crank Slot Mechanism
For more information, see Modeling Flexible Bodies in a Multibody Analysis (p. 5). The input file (p. 58)
CrankSlot_Flexible.inp (available on your product distribution media) is used to perform the
flexible body portion of the analysis.
The following figures show the FE representation of the flexible Rod2 part and a representation of the
entire model:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
54 of ANSYS, Inc. and its subsidiaries and affiliates.
Performing the Flexible Body Analysis
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 55
Example Multibody Analysis: Crank Slot Mechanism
Using CMS for static and transient nonlinear analysis reduces problem size and minimizes CPU-
resource requirements. You can convert parts of a model which exhibit linear behavior (such as
Rod2 in this case) to a superelement using CMS with large rotation. You can restrict all geometric,
contact, and material nonlinearity to those parts of the model which require nonlinear behavior.
For more information, see Using Component Mode Synthesis Superelements in a Multibody Ana-
lysis (p. 41) and Component Mode Synthesis in the Substructuring Analysis Guide.
Using the flexible body created previously (p. 54), create a component mode synthesis (CMS) model
with large rotation. Using CMS for the multibody analysis consists of:
3. Recovering stress and displacement results for the entire model (expansion pass).
To leverage the advantage of a CMS analysis for large rotation, define another part of the model, Rod1,
as a flexible body. Define the other flexible part, Rod2, as a CMS part. The input file (p. 58) Crank-
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
56 of ANSYS, Inc. and its subsidiaries and affiliates.
Using Component Mode Synthesis in the Multibody Analysis
Slot_FlexibleCMS.inp (available on your product distribution media) is used to perform the CMS
portion of the analysis.
The CMS part Rod2 assumes linear behavior with large rotations, whereas the flexible part Rod1 retains
all geometric and material nonlinearity in the model, as shown:
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 57
Example Multibody Analysis: Crank Slot Mechanism
CrankSlot_Rigid.inp
CrankSlot_Flexible.inp
CrankSlot_FlexibleCMS.inp
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
58 of ANSYS, Inc. and its subsidiaries and affiliates.
Chapter 7: Troubleshooting a Flexible Multibody Analysis
A successful flexible multibody simulation involves proper element selection, appropriate material be-
havior, and proper application of load and boundary conditions. To troubleshoot problems, debugging
must occur at all levels of the analysis. Typical questions requiring answers include:
• Is the choice of elements appropriate for this analysis? (For more information, see Element Choices for
Flexible Bodies (p. 6), Defining a Rigid Body (p. 6), and Connecting Multibody Components with Joint
Elements (p. 13).)
• Does the chosen material model correctly represent the actual material behavior?
• Are the loading and boundary conditions (p. 11) appropriately modeled?
• Do the problem's physics indicate global or local buckling issues that must be addressed?
Although other topics (p. 1) in this document provide guidelines for element selection, modeling, and
solver options while setting up your multibody analysis, the following troubleshooting topics are
available to help you achieve a successful multibody simulation:
7.1. Addressing Overconstraint Issues During Modeling
7.2. Resolving Overconstraint Problems
Mechanical APDL cannot always detect overconstraints automatically, particularly when the Lag-
range multiplier method is used. You are responsible for ensuring that the model is not overcon-
strained. Overconstrained models most often result in nonconvergence of the solution with small
solver pivot warnings, and in some cases may yield incorrect results. It is vital that you exercise
care when setting up your multibody simulation model.
Overconstraint means that more constraints than necessary have been applied to the degrees of freedom
(DOFs) at a node.
• Imposing boundary conditions on the DOFs at a given node if they are constrained via the CE or CP command.
• Contact modeling using the Lagrange multiplier method with improper boundary conditions on the contact
nodes.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 59
Troubleshooting a Flexible Multibody Analysis
The following examples illustrate scenarios in which overconstraint conditions can occur.
Consider the standard 3-D four-bar mechanism shown here. (See Geradin and Cardona in Learning More
About Multibody Dynamics (p. 3).) The mechanism consists of four rigid links and four revolute joints.
Revolute
x x
Joint
Pilot
node
With six DOFs available for each rigid body, the four rigid bodies yield a total of 6 * 4 = 24 DOFs. A re-
volute joint has only one free DOF and five constraints. Thus, the four revolute joints impose a total of
5 * 4 = 20 constraints. If one of the rigid links is fixed in space, then an additional six constraints are
imposed. If a rotation is applied at one of the revolute joints (thereby adding one more constraint), the
number of overconstraints is 24 - (20 + 6 + 1) = -3. As modeled, therefore, this mechanism is overcon-
strained.
In such a case, you can resolve the overconstraints by replacing three of the revolute joints with
spherical joints. Each spherical joint imposes only three constraints; after replacing the joint type, a DOF
count indicates that the system is no longer overconstrained. While the overconstraint in this model
can be resolved fairly easily, this is not a typical case. It is therefore vital that you exercise care when
setting up your model. For more information, see Resolving Overconstraint Problems (p. 62).
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
60 of ANSYS, Inc. and its subsidiaries and affiliates.
Addressing Overconstraint Issues During Modeling
A B
0 C
The figure shows a plate modeled with shell elements. A portion of the plate is made rigid by adding
MPC184 Rigid Beam elements (represented by the thick lines in the figure). The addition of rigid beams
AB and BC is redundant and leads to an overconstrained model.
In Mechanical APDL, if the MPC184 Rigid Beam elements with direct elimination option are used to
model this type of problem, the redundant constraints are eliminated automatically. However, if MPC184
Rigid Beam with the Lagrange multiplier option is used, the solution may not converge.
In some cases involving MPC184 Rigid Beam elements with the direct elimination option (which is based
on all DOFs at a node), redundant boundary conditions can result in an overconstrained system.
Consider a cylindrical tube with one end fixed and subjected to a bending moment at the other end.
A quarter of the cylinder is modeled with appropriate symmetry and antisymmetry boundary conditions
as shown in the following figure. MPC184 Rigid Beam elements with the direct elimination option
connect all the nodes of the tube to a center point, and a moment is applied at the center node.
Figure 7.3: Overconstrained System: Cylindrical Tube Subjected to Bending at One End
Because of the symmetry and antisymmetry boundary conditions, the system of internal constraint
equations generated due to the MPC184 Rigid Beam element results in an overconstrained system.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 61
Troubleshooting a Flexible Multibody Analysis
Mechanical APDL does not resolve overconstraints automatically. To check for overconstraints, model
the multibody mechanism as a rigid mechanism using a rigid body solver.
• Perform a DOF count in the mechanism. Various methods are available for evaluating the number of free
DOFs in a given rigid body mechanism. See Learning More About Multibody Dynamics (p. 3).
• Know the number of constraints for each joint element. In some cases, replacing one type of joint with an-
other may resolve an overconstraint issue. Check the number of constraints for a given joint and replace it
with a simpler one if possible. For example, a revolute joint (which imposes five constraints) can possibly
be replaced by a cylindrical joint (which imposes only four constraints). For more information, see Joint
Element Types (p. 13).
• A translational joint fixes five DOFs while allowing motion in only one direction. You may be able to replace
it with a slot joint which allows more free relative DOFs.
• The local axes specified at the joint element nodes must be defined properly. Improper definitions result in
unanticipated motion or constraints. For example, if you define the four-bar mechanism in Figure 7.1: Over-
constrained System: Standard 3-D Four-Bar Mechanism (p. 60)in a plane other than one of the global
Cartesian planes, verify that the joint coordinate systems for each joint align.
• Perform a modal analysis to ensure that appropriate modes are present in the idealized model of the
mechanism. Overconstraints can lead to modes that are not usually present in the actual system.
• Use more flexible components (p. 5) in the model. Avoid models with only rigid bodies (p. 6), which can
lead to solver difficulties.
• Avoid external (user-defined) constraint equations (CE and CP). They may conflict with those generated in-
ternally by Mechanical APDL for contact with MPC and the joint elements.
• Do not mix MPC184 Rigid Beam/Link and MPC184 Joint elements implemented using the Lagrange multi-
plier method with those implemented using the direct elimination method.
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
62 of ANSYS, Inc. and its subsidiaries and affiliates.
Index
M
multibody analysis
additional sources of information, 3
ANSYS-ADAMS interface, 3
boundary conditions for rigid bodies, 11
complex model representation using rigid bodies,
12
connecting bodies to joints, 26
connecting flexible and/or rigid components, 13
connecting joint elements to rigid bodies, 12
convergence criteria, 31
damping methods, 35
defining a rigid body, 6
definition, 31
element choices for flexible bodies, 6
energy output, 40
example analysis: crank slot mechanism, 51
finite element method benefits, 1
flexible body modeling, 5
initial conditions, 32
introduction, 1
joint element types, 13
kinematic constraints, 31
material behavior in joint elements, 20
modeling contact with rigid bodies, 13
modeling criteria, 5
overconstraint problems, 59
POST1 results, 37
POST26 results, 38
process overview, 2
results viewing, 37
rigid body DOFs, 10
rigid body modeling, 6
SMISC quantities for joint elements, 39
solver options, 36
time stepping, 36
troubleshooting, 59
using CMS superelements, 41
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates. 63
Release 18.2 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
64 of ANSYS, Inc. and its subsidiaries and affiliates.